{"resourceId":"nascio-2026-state-cio-survey-agentic-genai-20260928","versions":[{"version":"external-07ca7887016c823d3511a8b3da4af7f14eeefbc68d4844e328f669b3f9499653","resource":{"id":"nascio-2026-state-cio-survey-agentic-genai-20260928","title":"NASCIO 2026 State CIO Survey: 98% have AI enterprise policies; 64% say agentic AI will be most impactful","organization":"National Association of State Chief Information Officers (NASCIO)","sector":"State CIO association survey / enterprise IT governance","geography":"United States — 51 state and territory CIOs (summer 2026 survey)","publishedAt":"2026-09-28","publicationDate":"2026-09-28","eventDate":"2026-09-28","sourceName":"NASCIO","sourceLabel":"The 2026 NASCIO State CIO Survey — The State CIO in Motion: Priorities, Pressures and Progress","sourceUrl":"https://www.nascio.org/resource/the-2026-nascio-state-cio-survey-the-state-cio-in-motion-priorities-pressures-and-progress/","evidenceClass":"public-sector-association","outcomeClass":"emerging","topics":["governance-procurement","developers-agents","operating-model","knowledge-work"],"finding":"NASCIO's 17th annual State CIO Survey (released 2026-09-28 at the Annual Conference; 51 CIO respondents) reports AI as the top priority and near-universal enterprise AI policies (98%, up from 76% in 2025). GenAI procurement terms/contract provisions and transparency/accountability each rose to 61% (from 41%). For the first time the survey asked about agentic AI: 29% report current use, 35% planned, 29% no; 64% say agentic AI will be the most impactful emerging technology over 2–3 years versus 14% for GenAI. Common agentic uses are internal workflow/process automation, IT ops, coding/modernization and customer service, with CIOs stressing human accountability before judgment-sensitive automation.","sledRelevance":"Primary multi-state CIO signal for the state-government stream on GenAI procurement maturity and the shift toward agentic AI. Why now: released 2026-09-28 (NASCIO conference week) and not in the archive; directly matches the stream brief's NASCIO/agentic/procurement queries one day after Digital States leftovers filled recent editions.","evidence":"Opened the NASCIO resource HTML via curl (User-Agent SLED-research-bot/1.0) and extracted survey narrative text stating 98% enterprise policies, 61% GenAI procurement terms, first-time agentic question (29%/35%/29%), and 64% naming agentic AI most impactful. Corroborated release timing and key figures with same-day StateScoop and GovTech conference coverage without substituting those as the canonical URL.","architectureImplications":"Treat agentic pilots as internal workflow/IT-ops/coding agents with human approval gates before any rights-touching or procurement-judgment automation; inventory GenAI and agentic tools separately from chat copilots.","governanceImplications":"Near-universal policies do not equal complete inventories or mandatory standards—buyers should map which of the survey's GenAI action items (procurement terms, transparency, data governance) are actually enforced in contracts and agency SOPs.","securityPrivacyImplications":"Agentic tool-use expands blast radius beyond text generation; require least-privilege connectors, audit logs of tool calls, and data-governance rules called out by 61% of CIOs as still uneven.","caveats":"Self-reported CIO survey, not an independent outcome audit; agentic definitions may vary by respondent; full PDF may be member-gated though narrative stats appear on the public resource page; a secondary blog misstated current agentic use as 21% versus NASCIO/StateScoop/GovTech ~29%.","streamIds":["state-government"],"roles":{"sales":"Frame the NASCIO 2026 CIO Survey as the multi-state signal that GenAI policy is nearly universal while agentic AI is the next CIO bet—not as proof any single product works. Discovery should ask which of the survey's GenAI action items (enterprise policy, procurement terms, transparency, inventory) the state has actually enforced, and where agentic pilots sit (workflow vs eligibility). Bounded engagement: a 60–90 day agentic readiness assessment against the survey benchmarks and the state's contract riders. Unsupported claims include guaranteed hours saved, that 98% policy equals mature oversight, or that 29% agentic use implies production citizen agents.","engineering":"Fit is a gap analysis of GenAI/agentic controls versus NASCIO's reported practices, then a narrowly scoped internal agent (workflow, IT ops, or coding assist) with human approval. Architecture should separate retrieval/chat copilots from tool-calling agents, log every tool invocation, and keep procurement/eligibility decisions out of autonomy. Prerequisites: enterprise AI policy, inventory, and GenAI contract clauses. Constraints: survey self-report; definitions of agentic vary. Proof of value is a logged human-in-loop pilot with refusal/error rates—not CIO poll percentages.","delivery":"Delivery work includes mapping current GenAI inventory to the survey's action items, writing agentic SOPs with accountable owners, and negotiating procurement riders where terms are missing. Dependencies: CIO/CAIO, procurement, privacy, agency owners. Ownership of production agent behavior stays with the business owner, not the model vendor. Acceptance criteria: published inventory entry, contract AI clauses, and audited human approval on consequential actions. Risks: mistaking policy checkboxes for oversight, over-scoping agentic autonomy, and election-year CIO turnover disrupting pilots."},"roleDetails":{"sales":{"customerProblem":"State CIOs need a credible benchmark for GenAI maturity and whether to green-light agentic pilots amid election-year turnover.","stakeholders":["State CIO/CAIO","Procurement","Agency business owners","Privacy/security","Legislative oversight"],"discoveryQuestions":["Which GenAI action items from NASCIO (policy, procurement terms, inventory, transparency) are enforced today?","Where is agentic AI already running—internal workflow only or citizen-facing?","Who approves consequential agent actions?","What happens to AI programs if the CIO changes in 2027?"],"valueHypothesis":"Aligning pilots to NASCIO's GenAI/agentic benchmarks can de-risk investment if paired with contract terms and human gates—not poll headlines alone.","potentialEngagement":"60–90 day agentic/GenAI readiness assessment mapped to survey action items and current master contracts.","unsupportedClaims":["Guaranteed productivity savings from agentic AI","98% policy rate equals complete AI inventory","29% agentic use means citizen-facing autonomy is safe","Survey proves any specific vendor platform"]},"engineering":{"fit":"Internal agentic workflow/IT-ops/coding assist with mandatory human approval; not eligibility or rights adjudication.","architectureIntegration":"Enterprise IdP + tool registry + audit log of agent actions + human approval service before external side effects.","prerequisites":["Enterprise AI policy","Use-case inventory","GenAI contract clauses","Data classification for connectors"],"constraints":["Self-reported survey definitions","Election-year leadership churn","Procurement automation still early per survey"],"security":"Least-privilege tool credentials, prompt/tool-call logging, no training on restricted citizen data without agreement.","proofOfValue":["Measure task-completion with human override rate","Compare agent error/refusal rates to baseline staff workflow","Verify audit-log completeness for sampled actions"]},"delivery":{"work":"Inventory GenAI/agentic uses; draft agentic SOP with accountable owner; close procurement-term gaps; run one internal pilot.","dependencies":["CIO office","Procurement legal","Agency process owner","Security/privacy review"],"ownership":"Agency business owner accountable for agent outcomes; CIO sets enterprise guardrails.","skillsAdoption":"Train staff on when to trust agent drafts vs when human judgment is mandatory; short agent literacy modules.","governanceCheckpoints":["Pre-pilot risk review","Contract rider sign-off","Go-live inventory publication","30-day audit-log sample review"],"acceptanceCriteria":["Agent listed in inventory with owner","Human approval required for consequential actions","GenAI/agentic clauses present in governing contract"],"risks":["Policy theater without inventory","Autonomous actions in judgment workflows","CIO turnover stranding pilots"]}},"retrievedAt":"2026-09-29T22:20:20-05:00","enrichedAt":"2026-09-29T22:20:20-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Survey notes digital accessibility support to locals remains limited (StateScoop cites 35%); agentic/GenAI rollouts need workforce AI literacy and accessibility review of new agent UIs.","procurementImplications":"Use the 61% GenAI procurement-terms benchmark as a gap checklist—require AI disclosure, data-use limits, audit rights, and human-approval clauses before agentic or GenAI master-contract awards.","operatingModelImplications":"CIOs describe proving agentic value internally with human accountability first; operating model should keep legal/procurement judgment outside autonomous loops until evaluation templates exist.","sourceVerification":{"responseId":"curl-nascio-cio-survey-20260929T222020","openedUrl":"https://www.nascio.org/resource/the-2026-nascio-state-cio-survey-the-state-cio-in-motion-priorities-pressures-and-progress/","referenceExcerpt":"Sixty-four (64) percent said that agentic AI will be the most impactful emerging technology","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}